Thickness Trends and Sequence Stratigraphy

Home , Fossiliferous limestone, Hamilton Group, Marcellus Formation, Onondaga Limestone

Thickness trends and sequence AUTHORS Gary G. Lash Department of Geosciences, stratigraphy of the Middle State University of New York, Fredonia, Fredonia, New York 14063; [email protected] Devonian Marcellus Formation, Gary Lash received his B.S. degree from Kutztown State University (1976) and his M.S. degree Appalachian Basin: and Ph.D. from Lehigh University (1978 and 1980, respectively). He has worked on various as- Implications for Acadian pects of the Middle and Upper Devonian shale successions on western New York. Prior to this work, Lash was involved in stratigraphic and foreland basin evolution structural investigations of thrusted Cambrian– Ordovician deposits of the central Appalachians. Gary G. Lash and Terry Engelder Terry Engelder Department of Geosci- ences, Pennsylvania State University, University Park, Pennsylvania 16802; [email protected] ABSTRACT Terry Engelder, a leading authority on the Mar- Analysis of more than 900 wireline logs indicates that the cellus gas shale play, received his B.S. degree Middle Devonian Marcellus Formation encompasses two third- from the Pennsylvania State University, his M.S. degree from Yale University, and his Ph.D. from order transgressive-regressive (T-R) sequences, MSS1 and MSS2, Texas A&M University. He is currently a professor in ascending order. Compositional elements of the Marcellus of geosciences at Penn State and has previ- Formation crucial to the successful development of this emerg- ously served on the staff of the U.S. Geological ing shale gas play, including quartz, clay, carbonate, pyrite, and Survey, Texaco, and Lamont-Doherty Earth organic carbon, vary predictably within the proposed sequence- Observatory. He has written 150 research ar- stratigraphic framework. Thickness trends of Marcellus T-R ticles, many focused on Appalachia, and a book, Stress Regimes in the Lithosphere. sequences and lithostratigraphic units reflect the interplay of Acadian thrust-load-induced subsidence, short-term base-level fluctuations, and recurrent basement structures. Rapid thick- ACKNOWLEDGEMENTS ening of both T-R sequences, especially MSS2, toward the We thank AAPG reviewers Ashton Embry, northeastern region of the basin preserves a record of greater Berry Tew, Charles Boyer II, Ursula Hammes, accommodation space and proximity to clastic sources early and Karen Glaser for their efforts to improve in the Acadian orogeny. However, local variations in T-R se- this paper. Chief Oil and Gas, LLC, the New York quence thickness in the western, more distal, area of the basin State Energy Research and Development Au- thority, the Reservoir Characterization Group, may reflect the reactivation of inherited Eocambrian basement New York State Museum, the Ohio State Geo- structures, including the Rome trough and northwest-striking logical Survey, the Pennsylvania Geological Sur- cross-structural discontinuities, induced by Acadian plate con- vey, and Appalachian Fracture Systems sup- vergence. Episodes of block displacement locally warped the ported various aspects of this study. Randy Blood basin into northeast-southwest–trending regions of starved sedi- is acknowledged for the many field discussions mentation and/or erosion adjacent to depocenters in which re- related to the lithostratigraphy and sequence stratigraphy of the Middle and Upper Devonian gressive systems tract deposits were ponded. Block movement shale succession of the Appalachian Basin, including the Marcellus Formation. The AAPG Editor thanks the following reviewers for their work on this paper: Charles M. Boyer II, Ashton F. Embry, Karen S. Glaser, Ursula Copyright ©2011. The American Association of Petroleum Geologists. All rights reserved. Hammes, and Berry H. Tew Jr. Manuscript received September 4, 2009; provisional acceptance December 7, 2009; revised manuscript received March 22, 2010; final acceptance June 30, 2010. DOI:10.1306/06301009150

AAPG Bulletin, v. 95, no. 1 (January 2011), pp. 61–103 61 appears to have initiated in late Early Devonian Results reported on in this article are based on time, resulting first in thinning and local erosion of our analysis of more than 900 wireline logs from the Oriskany sandstone in northwest Pennsylvania. the Appalachian Basin of Pennsylvania, New York, This study, in addition to providing the basis for a northern West Virginia, eastern Ohio, and western predictive sequence-stratigraphic model that can Maryland (Figure 2). We focus on the Appalachian be used to further Marcellus exploration, tells of a Plateau region of the basin for three reasons: its foreland basin more tectonically complex than ac- greater density of available wireline logs, fewer counted for by simple flexural models. structural complications, and greater economic via- bility. Specific points addressedinthisstudyinclude (1) the distribution and thickness trends of the two black shale members of the Marcellus Formation; INTRODUCTION (2) the distribution and thickness of the intervening limestone, an interval that could be critical to stim- The much studied Hamilton Group of the Appala- ulation and production considerations; and (3) the chian Basin is an eastward and southeastward thick- stratigraphy and distribution of the organic-rich ening wedge of marine and nonmarine shale, silt- Levanna Shale Member of the Skaneateles Forma- stone, and sandstone (Cooper, 1933, 1934; Rickard, tion, a unit that has been confused with the Mar- 1989). To the west of the Hudson Valley, the Ham- cellus Formation. ilton succession is defined by an increasing abun- As important as the previous points are, how- dance of fossiliferous black and gray marine shale ever, the more significant contribution of this ar- interbedded with several thin laterally extensive ticle is a sequence-stratigraphic framework of the limestones (Cooper, 1930, 1933; de Witt et al., Marcellus Formation as deduced from publicly 1993). Hamilton Group deposits, part of the Cats- available wireline logs. Partington et al. (1993) and kill delta succession, accumulated in an elongate Emery and Myers (1996), among others, have foreland basin that formed in response to the Aca- demonstrated the use of some of the more com- dian oblique collision of the Avalonia microplate and mon wireline log suites to the interpretation of Laurentia (Figure 1; Ettensohn, 1985, 1987; Faill, sedimentary successions in terms of such sequence- 1985; Ferrill and Thomas, 1988; Rast and Skehan, stratigraphic elements as sequence boundaries, sys- 1993). Throughout much of the basin, the Middle tems tracts, condensed sections, and maximum Devonian Marcellus Formation (Hall, 1839), the flooding surfaces. Such an approach serves as a basal unit of the Hamilton Group, comprises two means by which basin fill can be organized into black shale intervals separated by a sequence of unconformity (or equivalent conformable surface) limestone, shale, and lesser sandstone of variable bounded packages of strata that provide a frame- thickness (Clarke, 1903; Cate, 1963; de Witt et al., work for predictive reservoir assessment and cor- 1993). Recent advances in well stimulation of, and relation into regions of minimal or poor data con- production from, Devonian–Mississippian black trol. Thickness trends of lithostratigraphic units and shale gas deposits of the eastern and southern the sequence stratigraphy of the Marcellus Forma- United States, including the Barnett, Fayetteville, tion reveal a basin that was more tectonically active and Woodford shales, have yielded similarly prom- than heretofore realized. Reactivated extensional ising results in the Marcellus Formation (Engelder basement structures, including Eocambrian faults et al., 2009). Continued successful exploration and associated with the Rome trough, and northwest- exploitation of the Marcellus necessitates an en- striking wrench faults (i.e., cross-strike structural hanced understanding of its stratigraphy, including discontinuities of Wheeler, 1980), appear to have thickness trends and compositional attributes of controlled sedimentation patterns of at least the the component units, throughout the core region upper Lower through lower Middle Devonian suc- of Marcellus production beneath the Appalachian cession, including the Marcellus Formation, in west- Plateau. ern New York and northwest Pennsylvania.

62 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 1. (A) General tectonic reconstruction of the Appalachian orogenic belt during the Acadian orogeny showing approximate locations of synorogenic deposits. Modified from Ettensohn (1992) and Ferrill and Thomas (1988). (B) Diagram illustrating the inferred relationship among thrust loading, foreland basin subsidence, black shale sedimentation, and forebulge development. Modified from Ettensohn, 1994.

Lash and Engelder 63 Figure 2. Base map of the core region of the Marcellus Formation basin in New York, Pennsylvania, eastern Ohio, western Maryland, and northern West Virginia. Symbols indicate the locations of wireline logs that were used in this study; numbered and lettered wells and cross-sections refer to text figure numbers. WC = Wayne County, Pennsylvania; SUC = Susquehanna County, Pennsylvania; WYC = Wyoming County, Pennsylvania; WAC = Washington County, Pennsylvania; GC = Greene County, Pennsylvania; SC = Sullivan County, New York; CC = Cayuga County, New York; BC = Broome County, New York.

Figure 3. Comparison of the Marcellus Formation stratigraphy used in this study with the recently revised Marcellus stratigraphic nomenclature of Ver Straeten and Brett (2006).

64 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 4. Wireline log of the Rodolfy well in Wayne County, Pennsylvania, illustrating (1) the Marcellus stratigraphy of Ver Straeten and Brett (2006), (2) the lithostratigraphy used in this study, and (3) the sequence stratigraphic interpretation of the log (refer to text for discussion). See Figure 2 for location of the Rodolfy well. RST = regressive systems tract; MFS = maximum flooding surface; TST = transgressive systems tract; MRS = maximum regressive surface.

Lash and Engelder 65 Figure 5. Isochore map of the Union Springs Member. The dashed isochore represents the mapped zero line of the Union Springs Member; elsewhere, contouring is limited by a lack of data. Isochore lines were not traced into the valley and ridge because of sparse data and structural complexities.

STRATIGRAPHIC FRAMEWORK shale exposed near the village of Marcellus, Onon- daga County, New York. Henry Darwin Rogers, a James Hall (1839) was the first to apply the name contemporary of Hall, referred to Marcellus equiv- “Marcellus shale” to organic-rich black and gray alents exposed in Pennsylvania as the “Cadent Lower

Figure 6. Wireline log signature of a very thin Union Springs Member as exhibited in the Stowell–Kolb well, Chautau- qua County, New York; M = Levanna Member of the Skan- eateles Formation; S = Stafford Member of the Skaneateles Formation.

66 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 7. Cross-section illustrating lateral variations in the thickness of the Union Springs Member and overlying Cherry Valley Member; OL = Onondaga Limestone; US = Union Springs Member; CV = Cherry Valley Member; OC = Oatka Creek Member; Depth = drilling depth.

Black and Ash-Colored Slate” in his report of the where the Marcellus Formation is part of a gen- First Pennsylvania Geological Survey started in erally shallowing-upward trend from basinal black 1836 (Millbrooke, 1981). Eighty years lapsed be- shale to nearshore sandstone and fluvial deposits. fore Cooper (1930) subdivided the Marcellus For- Specifically, Ver Straeten and Brett (2006) raised the mation into the Union Springs Member and over- Marcellus Formation to the subgroup level and the lying Oatka Creek Member. This nomenclature, or Union Springs and overlying Oatka Creek members some modification of it, largely stemming from out- to the formation level (Figure 3). Such a modifica- crop investigations in New York, has been adopted tion is consistent with the nomenclature of the by most workers who have studied the lower Ham- overlying units of the Hamilton Group, including ilton Group in the subsurface of New York, Penn- the Skaneateles Formation, Ludlowville Formation, sylvania, and Ohio (e.g., Oliver et al., 1969; Van and Moscow Formation, in ascending order. Tyne, 1983; Rickard, 1984, 1989). The newly defined Union Springs Formation In a series of articles spanning nearly 15 years, is principally composed of black shale of the Ba- Ver Straeten et al. (1994), Ver Straeten and Brett koven Member (Ver Straeten and Brett, 2006). The (1995, 2006), and Ver Straeten (2007) proposed a organic-rich Bakoven Member of eastern New York Marcellus stratigraphy that seeks to reduce the ac- is overlain by the Stony Hollow Member of the cumulated, sometimes confusing, stratigraphic ver- Union Springs Formation (Figure 3), a succession of biage of more than 150 yr of study. The revised calcareous shale, siltstone, and fine-grained sand- stratigraphy links the generally fine-grained Mar- stone. In western New York, fossiliferous limestone cellus succession of the more distal western region and dark shale of the Hurley Member of the Oatka of the basin with that of the proximal eastern basin Creek Formation overlies the Bakoven Member

Lash and Engelder 67 Figure 8. Wireline logs illustrating the gamma-ray signatures of the Union Springs and overlying Oatka Creek members of the Marcellus Formation: (A) Gradational contact of the Cherry Valley and overlying Oatka Creek Member; (B) Union Springs–Oatka Creek contact (Cherry Valley Member is absent).

(Figure 3) (Ver Straeten and Brett, 2006). To the and Ver Straeten, 1995) (Figure 3). The Cherry east, the Hurley Member of the Oatka Creek– Valley is sharply overlain by black and gray shale of equivalent Mount Marion Formation overlies the the Berne Member of the Oatka Creek and Mount Stony Hollow Member (Figure 3). The Hurley Marion formations (Figure 3). The Berne Member Member is overlain by bedded and nodular fine- of the more proximal eastern region of the basin grained limestone of the Cherry Valley Member passes upward into organic-lean strata of the Otsego, (Ver Straeten et al., 1994; Ver Straeten and Brett, Solsville, and Pecksport members, in ascending order 1995, 2006). The latter thickens to the east and (Ver Straeten and Brett, 1995). The top of the Mar- south to as much as 32 ft (10 m) south of Albany, cellus subgroup is defined by the base of the calcar- New York, where it is composed of interbedded eous Stafford Member of the Skaneateles Formation shale and bioturbated sandstone (Ver Straeten and in western New York and northwestern Pennsylva- Brett, 1995). The Purcell Member of Pennsylvania nia and the laterally equivalent Mottville Member is equivalent to the Cherry Valley Member (Brett of central New York (Ver Straeten and Brett, 1995).

68 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin 4). Unfortunately, the paucity of wireline logs from this area of the basin precludes the more wide- spread application of this stratigraphy. In this article, we adopt a lithostratigraphy more in line with that used by Rickard (1984, 1989) and one that lends itself to subsurface correlation of wireline log signatures. Specifically, we define our basal unit of the Marcellus Formation as the Union Springs Member, a term recognized by the U.S. Geological Survey Geologic Names Lexicon (USGS, 2008). The Union Springs Member of this study, which encompasses the Bakoven Member of Ver Straeten and Brett (2006), is overlain by the Cherry Figure 9. Interbedded black shale and limestone defining the Valley Member (Figure 3). Our Cherry Valley gradational contact of the Onondaga Formation and overlying Member, which is composed of variable amounts Union Springs Member of the Marcellus Formation exposed of interlayered carbonate, shale, and sandstone, along the Norfolk Southern mainline, Hamilton-Newton, Penn- correlates with the Stony Hollow Member of the sylvania. Hammer = 30 cm (11.8 in.) long. Union Springs Formation and the Hurley and Cherry Valley members of the Oatka Creek and Mount Marion formations of Ver Straeten and The revised Marcellus stratigraphy is a welcome Brett (2006) (Figure 3). Finally, we use the name contribution to our understanding of Appalachian Oatka Creek Member, also recognized by the Geo- Basin stratigraphy. However, such a detailed level logic Names Lexicon, for the succession of black of subdivision, especially its differentiation of grossly and gray shale and lesser siltstone and limestone similar units such as the Hurley and Cherry Valley that underlies the Stafford and Mottville members members is not readily applicable to basinwide of the Skaneateles Formation (Figure 3). stratigraphic interpretation of publicly available wireline logs of varying quality. Indeed, only one log available to us permits a tentative identification of the Marcellus stratigraphy espoused by Ver Straeten MARCELLUS FORMATION: SUBSURFACE and Brett (2006). The Marcellus Formation dis- LITHOSTRATIGRAPHY AND THICKNESS TRENDS played by the Columbia 1 Rodolfy well of Wayne County, Pennsylvania, includes upper and lower The subsurface stratigraphy of the Marcellus For- black shale intervals separated by more than 110 ft mation has been addressed by a number of workers. (34 m) of calcareous and arenaceous rock (Figure 4). Harper and Piotrowski (1978, 1979) and Piotrowski Gamma-ray and neutron porosity signatures per- and Harper (1979) published isopach maps of De- mit the subdivision of the middle horizon into vonian units in Pennsylvania, including the Mar- upper and lower intervals (Figure 4). The latter is cellus Formation. Their maps present the net thick- composed of approximately 65 ft (20 m) of what ness of the radioactive facies of the Hamilton Group, appears to be sandy limestone, the likely equiv- principally the Marcellus Formation. Rickard (1984, alent of Ver Straeten and Brett’s (2006) Stony 1989) mapped the Marcellus in the subsurface of Hollow Member of their Union Springs Formation New York, northeastern Ohio, and the northern two (Figures 3, 4). These deposits are overlain by ap- thirds of Pennsylvania. He presented isopach maps proximately 45 ft (14 m) of poorly radioactive low- for what he termed the lower Marcellus Formation porosity strata that include the Hurley and over- (Union Springs Member–Cherry Valley Member lying Cherry Valley members of Ver Straeten and interval), the upper Marcellus Formation (Oatka Brett’s (2006) Mount Marion Formation (Figures 3, Creek Member and equivalent units), and the total

Lash and Engelder 69 Figure 10. Variations in the Onondaga Limestone–Marcellus Formation contact across the Appalachian Basin; (A–C) gradational Onondaga–Union Springs contact; (D) sharp Onondaga-Union Springs contact; (E–G) = sharp Onondaga-Oatka Creek contact. Note that the wireline log D is a gamma-ray log only.

Marcellus Formation. Last, a series of stratigraphic Union Springs Member cross-sections and isopach maps of the Devonian clastic succession of the Appalachian Basin, includ- The Union Springs Member of this report thickens ing the Marcellus Formation, were published as to the east and southeast from western New York; part of the Eastern Gas Shales Project of the late it is especially thick in northeastern Pennsylvania 1970s and early 1980s (e.g., de Witt et al., 1975; where it exceeds 160 ft (49 m) (Figure 5). Particu- West, 1978; Roen et al., 1978; Kamakaris and Van larly intriguing, though, is the local absence of the Tyne, 1980). Union Springs along a northeast-southwest–trending

70 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 10. Continued. axis in western New York into northwestern Penn- and total organic carbon (TOC) content. Thin sylvania (Figure 5). Examination of close-spaced carbonate (concretionary?) intervals and pyrite-rich wireline logs from this area of this basin reveals the layers can be recognized on gamma-ray, bulk den- occasional presence of the Union Springs Member sity, and photoelectric index wireline logs. The up- as a thin, radioactive, low-density interval imme- per three quarters or so of the Union Springs is diately above the Onondaga Formation and below defined by a generally diminished gamma-ray re- the Cherry Valley Member (e.g., Rickard, 1984) sponse (Figure 8) and increased bulk density. (Figures 6, 7). Although the Tioga Ash Bed is rec- The contact of the Onondaga Formation and ognized by sharp increases in gamma-ray response, overlying Union Springs Member has been inter- bulk density signatures of these deposits, which preted to be a regional unconformity (Potter et al., remain close to the gray shale level, enable one to 1982; Rickard, 1984, 1989). However, Ver Straeten differentiate ash layers from thin organic-rich in- (2007) maintains that with the exception of central tervals of the Union Springs Member. New York through eastern Pennsylvania, the con- The basal 20 to 30 ft (6–9 m) of the Union tact is relatively conformable across much of Penn- Springs Member is especially radioactive (locally sylvania, western New York, Ohio, Maryland, and >600 American Petroleum Institute [API] units) West Virginia (Figure 9). The absence of the Union and low density (<2.35 g/mL). Proprietary data dis- Springs Member from western New York and north- cussed below reveal this interval to be characterized western Pennsylvania complicates this question. by reduced clay and markedly higher quartz, pyrite, Clearly, in those regions of the basin lacking the

Lash and Engelder 71 Figure 11. Isochore map of the Cherry Valley Member. The dashed isochore represents the mapped zero line of the Cherry Valley Member; elsewhere, contouring is limited by a lack of data. Isochore lines were not traced into the valley and ridge because of sparse data and structural complexities.

Union Springs Member (Figure 4), the Onondaga– Cherry Valley Member Marcellus contact is unconformable. The Onondaga– Union Springs contact appears to cut progressively The Union Springs Member of western New York deeper into the Onondaga Limestone from central is overlain by the Cherry Valley Member, the cor- New York eastward to the Hudson Valley (Rickard, relative of the Purcell Limestone of Pennsylvania 1989). Oliver (1954, 1956), however, described and West Virginia, an interval of bedded and nod- the exposed contact in Cayuga County, New York, ular limestone, shale, and siltstone (Cate, 1963; as an approximately 7-ft (2.1 m)-thick interval of Dennison and Hasson, 1976). As noted previously, interbedded shale and limestone, suggestive of a the Cherry Valley Member of this report includes more gradational contact. Moreover, gamma-ray the interval encompassing the Stony Hollow Mem- and bulk density logs reveal a seemingly gradational ber of the Union Springs Formation of Ver Straeten contact extending across central and eastern New and Brett (2006) and the Hurley and Cherry Valley York and south through Pennsylvania into north- members of Ver Straeten and Brett’s (2006) Mount ern West Virginia (Figure 10A, B, C). However, Marion Formation (Oatka Creek Formation equiv- the Onondaga–Union Springs contact of the West alent) of the more proximal eastern region of the Virginia northern panhandle and eastern Ohio is basin (Figure 2). very sharp, although not necessarily unconform- The Cherry Valley Member increases from less able (Figure 10D). than 10 ft (3 m) thick in western New York and

72 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 12. Isochore map of the Oatka Creek Member. The dashed isochore represents the mapped zero line of the Oatka Creek Member; elsewhere, contouring is limited by a lack of data. Isochore lines were not traced into the valley and ridge because of sparse data and structural complexities.

northwestern Pennsylvania to more than 140 ft Although present throughout much of the ba- (43 m) thick in Wayne County, northeastern Penn- sin, the Cherry Valley Member displays rather ir- sylvania, and Sullivan County, southeastern New regular thickness trends (Figure 11), similar to what York, as well as northeastern West Virginia (Figure 11). can be observed at the outcrop scale (e.g., Ver The rate of change of thickness is especially rapid Straeten et al., 1994). Note that the Cherry Valley in southeastern New York, a likely reflection of the is absent along a northeast-southwest–trending re- presence of Ver Straeten and Brett’s (2006) Stony gion of western New York and northwestern Penn- Hollow Member (Figures 4, 11). Indeed, gamma- sylvania, coincident with that area of the basin from ray and bulk density log signatures indicate that which the Union Springs Member is thin or absent the Cherry Valley interval becomes more arena- (compare Figures 5, 11). Locally, the Cherry Val- ceous to the east (see Figure 4), an observation ley Member overlies the Onondaga Limestone, the consistent with field relationships documented from intervening Union Springs Member absent because southeastern New York (Ver Straeten et al., 1994). of erosion or nondeposition (Figure 7). In this case, Moreover, an ongoing coring program to sample the Cherry Valley is recognized as a plateau on the the Marcellus Formation in the Pennsylvania Val- gamma-ray signature immediately beneath the ra- ley and Ridge confirms the presence of an arena- dioactive basal interval of the overlying Oatka Creek ceous Purcell Limestone in this region of the basin. Member (Figure 7).

Lash and Engelder 73 Oatka Creek Member

TheCherryValleyMemberisoverlainbytheOatka Creek Member, recognized by a radioactive basal

nterval overlain by interval that passes upward into a higher density, less radioactive (lower TOC) shale succession con- taining occasional carbonate layers (Figure 8A). The organic-rich basal interval of the Oatka Creek is generally less radioactive than the basal interval of the Union Springs Member (Figure 8A, B). The contact of the Oatka Creek and underlying Cherry Valley is readily defined in wireline logs; locally, the contact appears to be gradational (Figure 8A). However, in the absence of the Cherry Valley (or perhaps a limestone interval thinner than tool resolution), the contact is placed a short distance below the gamma-ray peak in the radioactive basal interval of the Oatka Creek Member (Figure 8B). The Oatka Creek rests disconformably on the Onondaga Limestone in those areas of the basin absent in the Union Springs and Cherry Valley members. Indeed, in almost every example studied, well-log signatures indicate a very sharp contact (Figure 10E, F, G), quite unlike the seemingly gradational Onondaga Limestone–Union Springs Member contact described from much of the basin. The Oatka Creek Member, present throughout the core region of the basin, thickens to the east, most rapidly along a north-south line east of the meridian that defines the western edge of Broome County, New York, and the western boundaries of Susquehanna and Wyoming counties, Pennsylvania (Figure 12). It exceeds 550 ft (168 m) thick in eastern Wayne County, Pennsylvania, into Sullivan County, New York (Figure 12). However, thickening of the Oatka Creek Member is manifested principally by a marked increase in the thickness of the organic-lean upper part of the unit (Figure 13). That is, the basal radioactive interval displays an eastward increase in thickness, but at a rate far less than that of the overlying higher density (organic- lean) shale. TheOatkaCreekMemberthinstolessthan

Cross-section illustrating eastward thickening of the Oatka Creek Member. The Oatka Creek is subdivided into an30 organic-rich (radioactive) basal i ft (9 m) along a northeast-southwest–oriented axis in western New York, extending into Pennsyl- vania (Figure 12). This area is displaced to the east of the similarly oriented region of the basin over organic-lean shale (low radioactivity). The cross-section is hung on the top of the Oatka Creek Member. Figure 13.

74 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 14. ContactoftheOatka Creek Member, Marcellus For- mation, and the overlying Stafford Member of the Skaneateles For- mation exposed on Oatka Creek, village of LeRoy, New York. TOC of the Oatka Creek Member immediately beneath the contact is 4%. Note heavily jointed nature of the organic-rich shale.

which the Union Springs and Cherry Valley mem- carbonaceous shale (Figure 18). The subsurface bers are absent (compare Figures 5 and 11 with 12). Levanna Member, unlike the Union Springs and Thinning of the Oatka Creek Member is confined Oatka Creek members of the Marcellus Formation, to the organic-lean shale interval; that is, there is is recognized by increasing gamma-ray response no concomitant thinning of the radioactive low- upsection (Figure 19A). We arbitrarily place the density basal interval of the Oatka Creek across top of the Levanna at the stratigraphically highest this structure (Figure 14). Indeed, the organic-rich gamma-ray peak equal to or greater than 20 API facies of the Oatka Creek Member thickens across units in excess of a gray shale baseline established the region over which the unit thins (Figure 15). in the overlying nonsource succession of the Skan- eateles Formation (Figure 19A), an approach similar Stafford and Levanna Members, to that adopted for use by the Eastern Gas Shales Skaneateles Formation Project (de Witt et al., 1993). The most organic-rich facies of the Levanna Member passes laterally into The upper contact of the Oatka Creek Member undifferentiated organic-lean shale of the Skaneateles is defined by the base of the Stafford (Mottville) Formation (Figure 19B). Thus, although Levanna Member, an easily recognized marker in the sub- outcrops can be observed outside our isochore re- surface at the base of the Skaneateles Formation gion, it is likely that these deposits represent the (Oliver et al., 1969; de Witt et al., 1993) (Figure 16A). organic-lean lateral equivalent of the Levanna Mem- The bulk of the Stafford is present in a northeast- ber mapped in the subsurface. The bulk of the southwest–oriented region of western New York Levanna occupies a generally rectangular region and northwestern Pennsylvania; however, outliers of the basin extending from western New York to of the Stafford are present in southwestern Penn- southwestern Pennsylvania, thinning rapidly to the sylvania and western Maryland (Figure 17). The east and southeast, and more gradually to the west limestone appears to pass laterally into shale. and southwest (Figure 18). The Stafford Member is overlain by the Le- Although not part of the Marcellus Formation, vanna Member of the Skaneateles Formation, a the Levanna Member supplements the Middle De- relatively geographically restricted little-described vonian source rock inventory where it is present.

Lash and Engelder 75 76 tairpyo h ideDvna aclu omto,AplcinBasin Appalachian Formation, Marcellus Devonian Middle the of Stratigraphy

Figure 15. Interpretive fill cross-section of gamma-ray values. Note onlapping of organic-lean deposits on both sides of the central region of thinning. Indeed, in those regions of the basin lacking the Embry and Johannessen (1992) and further refined Stafford Member, the Levanna can be mistaken for by Embry (1993, 1995, 2002). Indeed, Johnson et al. the Marcellus Formation (de Witt et al., 1993). The (1985) first applied the T-R sequence concept to organic-rich facies of the Levanna Member appears the Devonian succession of the Appalachian Basin a to pass rapidly eastward into organic-lean low- quarter of a century ago. A single T-R sequence radioactivity shale (Rickard, 1989); the Levanna comprises a transgressive systems tract, a deepening- thins westward into Ohio where it aids in the rec- up succession that records rising base level, overlain ognition of the top of the Oatka Creek. In the ab- by regressive systems tract deposits that accumu- sence of the Stafford Member, the Oatka Creek lated during falling base level and consequent re- Member–Skaneateles Formation contact is placed duced accommodation space (Embry and Johan- at a density maximum and/or gamma-ray minimum nessen, 1992; Embry, 1993, 2002). Delimiting T-R (Figure 16B) or a subtle reduction (∼10–15 API sequences requires the identification of minimally units) in gamma-ray signature (Figure 16C)that diachronous sequence-boundary surfaces (Embry, can be traced laterally into the Stafford Member. 2002; Mancini and Puckett, 2002). Embry (2002) demonstrated that those surfaces most conducive to defining T-R sequences include the subaerial un- SEQUENCE-STRATIGRAPHIC FRAMEWORK conformity, the unconformable shoreline ravine- OF THE MARCELLUS FORMATION ment, and the maximum regressive surface. The previously mentioned surfaces are best Mapping of lithostratigraphic units in the subsur- thought of in terms of a single asymmetric base-level face is necessary to basin analysis and exploration cycle initiated by a rapid rise in base level (Embry considerations. However, it is the application of et al., 2007). Early in this base-level cycle, the sedi- the sequence-stratigraphic approach that enables ment flux to the shoreline may be high enough so one to subdivide basin fill into a framework of that the shoreline continues to advance toward the systems tracts and bounding and internal surfaces basin. Soon, however, the rate of base level rise at based on depositional models and asymmetric (ac- the shoreline exceeds the rate of sediment supply, tualistic) base level curves. The resulting framework resulting in a landward or transgressive movement of reservoir properties, including compositional at- of the shoreline. The maximum regressive surface tributes, can be used to reduce risk in frontier re- marks this change in depositional regime from regions of the basin or areas of poor data control. The gression to transgression (Embry, 2002; Embry use of wireline logs to sequence-stratigraphic anal- et al., 2007). Consideration of asymmetric base level ysis has been demonstrated by a number of authors, curves defined by a rapid initial rise in base level including Van Wagoner et al. (1990), Embry and suggests a coincidence of the maximum regressive Johannessen (1992), Embry (1993, 2002), Partington surface and the onset of the rise in base level (Embry et al. (1993), Emery and Myers (1996), Brown et al. et al., 2007). (2005), and Singh et al. (2008). Moreover, combi- Rapid trangresssion is accompanied by wave- nations of wireline logs can be used to interpret rock induced erosion, much of the sediment being trans- properties, including organic richness, in terms of a ported seaward. The resulting scoured surface, a sequence-stratigraphic framework (e.g., Passey et al., shoreline ravinement, may or may not have incised 1990; Creaney and Passey, 1993). through an underlying subaerial unconformity produced during the preceding regression. In the for- T-R Sequences: Background mer case, the shoreline ravinement is termed an unconformable shoreline ravinement and is overlain Sequence stratigraphy seeks to define and corre- by a deepening-upward marine succession, the late changes in depositional dynamics that reflect a transgressive systems tract, which reflects a reduced single base level cycle. This article adopts the T-R supply of sediment to the marine environment as sequence described by Johnson et al. (1985) and the shoreline migrates landward. The presence of

Lash and Engelder 77 Figure 16. Wireline logs illustrating the nature of the top of the Oatka Creek Member– Skaneateles Formation contact in the presence (A) and absence (B) of the Stafford Member. Panel (C) illustrates the approach taken in this study to define the contact in more proximal regions of the basin.

deepening-upward marine deposits immediately As the rate of base level rise slows to such a above both the unconformable shoreline ravine- point that the sediment flux at the shoreline ex- ment and the maximum regressive surface suggests ceeds the rate of sediment removal by wave action, that these surfaces are minimally diachronous lat- ravinement development stops and the shoreline eral equivalents, the former along the flanks of the begins a regressive seaward migration. Increased basin and the latter in the more basinal conform- sediment supply results in progradation of coarser able succession (Embry, 2002; Embry et al., 2007). sediments across the shelf thereby replacing the

78 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 16. Continued.

deepening-upward trend with a shallowing-upward or pass distally into, a condensed section (Emery succession that comprises the regressive systems and Myers, 1996; Jervey, 1988; Van Wagoner et al., tract. The surface defining the change from deep- 1988, 1990; Partington et al., 1993). Condensed ening-upward strata, the transgressive systems tract, sections are aerially extensive and can be composed to shallowing-upward deposits of the regressive sys- of abundant organic matter (high TOC) and au- tems tract is the maximum flooding surface (Embry, thigenic/diagenetic minerals (Loutit et al., 1988; 2002; Embry et al., 2007). The maximum flood- Posamentier et al., 1988; Sarg, 1988; Liro et al., ing surface does not necessarily record the base 1994; Emery and Myers, 1996). Abundant plank- level maximum. Indeed, regression can be initiated tonic fossils common to condensed sections (Loutit during a base level rise if the rate of rise is less than et al., 1988; Posamentier et al., 1988) reflect the the rate of sediment supply as might be expected of markedly reduced supply of clastic detritus to more tectonically driven base level change (Embry et al., basinal environments (Partington et al., 1993). 2007). Some have argued that transgressive systems Increased accommodation space and the re- tract deposits, especially the condensed section, have duced clastic sediment flux that attends transgres- the greatest source rock potential (e.g., Vail, 1987; sion during rising base level culminates in the maxi- Emery and Myers, 1996). Indeed, the reduced mum flooding surface, which may correspond with, clastic flux that attends a base level high favors the

Lash and Engelder 79 Figure 17. Isochore map of the Stafford Member of the Skaneateles Formation. The dashed isochore represents the mapped zero line of the Stafford Member; elsewhere, contouring is limited by a lack of data. Isochore lines were not traced into the valley and ridge because of sparse data and structural complexities.

concentration of oil-prone marine organic matter The general relationship of condensed section (Creaney and Passey, 1993). Moreover, the land- deposits and source rock potential, although con- ward shift of environments accompanying a rise firmed by a number of studies (e.g., Lambert, 1993), in base level induces the shoreward translation of is somewhat more complex than described (e.g., organic-rich facies, resulting in accumulation of the Mancini et al., 1993). Palsey et al. (1991) dem- condensed section close to or at the top of the onstrated that the greatest source potential interval transgressive systems tract. The described facies in the Upper Cretaceous Tocito Sandstone and shift is manifested by a general increase in TOC associated Mancos Shale of the San Juan Basin, upward from the base of the transgressive systems New Mexico, is not found in the condensed section tract (Creaney and Passey, 1993). Still, organic-rich within the shale, but rather 82 ft (25 m) below, in source rocks may continue to accumulate as part the Tocito Sandstone. Similarly, Leckie et al. (1990) of the regressive systems tract as relative base level demonstrated that the highest source rock poten- begins to drop. Further reduction of base level and tial in a succession of shallow water Cretaceous accommodation space, however, is normally ac- deposits in western Canada exists within the trans- companied by an increase in clastic sediment flux gressive systems tract, but below the condensed and consequent dilution of TOC (Creaney and section. However, Curiale et al. (1991) observed Passey, 1993). that the most promising source rock interval in the

80 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 18. Isochore map of the Levanna Member of the Skaneateles Formation. The dashed isochore represents the mapped zero line of the Levanna Member; elsewhere, contouring is limited by a lack of data. Isochore lines were not traced into the valley and ridge because of sparse data and structural complexities.

Cenomanian–Turonian succession of New Mexico that can be used for basinwide correlation. Indeed, lies above the condensed interval. Palsey et al. our composite wireline log serves as a type section (1991) suggested that in some cases, sedimenta- of the Marcellus Formation in the subsurface. The tion rates within those environments favorable to Marcellus encompasses the bulk of two T-R se- accumulation of condensed section deposits are quences herein referred to as MSS1 and MSS2, in too low to preserve abundant hydrogen-rich or- ascending order (Figure 20). These sequences, ap- ganic matter. proximate equivalents of Johnson et al.’s (1985) T-R cycles Id and Ie and Ver Straeten’s (2007) Eif-2 and Eif-3 sequences, span approximately 1.8 m.y., ex- Discussion tending from the upper “costatus” conodont zone through the “hemiansat” zone (Kaufmann, 2006; Our sequence-stratigraphic analysis of the Mar- Ver Straeten, 2007). The relatively short duration cellus Formation begins with the generation of a of MSS1 and MSS2 is consistent with their reflect- composite wireline log as described by Brown et al. ing third-order base level cycles (Mitchum and (2005) (Figure 20). Referred to as a site-specific Van Wagoner, 1991) that occurred within a second- sequence-stratigraphic section (S5) benchmark chart order cycle, encompassing much of the Middle and (Brown et al., 2005), the composite log places the Upper Devonian succession (Johnson et al., 1985). Marcellus into a sequence-stratigraphic framework Embry (1995) advocated a hierarchical system of

Lash and Engelder 81 Figure 19. Example of the approach used in this study to determine the thickness of the Levanna Member; (A) gamma-ray log dis- playing the Levanna Mem- ber and (B) gamma-ray log from a well lacking the organic-rich Levanna Member. Refer to text for discussion.

82 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 20. Sequence-stratigraphic type section of the Marcellus Formation that encompasses the upper part of the underlying Onondaga Formation and the lower interval of the Skaneateles Formation. TST = transgressive systems tract; RST = regressive systems tract; MFS = maximum flooding surface; MRS = maximum regressive surface. Refer to text for discussion.

T-R sequences based on sequence boundary charac- flanks are consistent with third-order T-R sequences teristics rather than duration. The generally basin- (Embry, 1995). wide extent of Marcellus T-R sequence boundaries, The maximum regressive surface that defines manifestly transgressive deposits overlying sequence the base of T-R sequence MSS1 is placed at a gamma- boundaries, and the presence of unconformities (un- ray minimum and/or bulk density maximum close to conformable shoreline ravinement) only on basin or at the top of the Onondaga Limestone (Figure 20).

Lash and Engelder 83 Figure 21. Mineralogy and total organic carbon (TOC) trends through the MSS1 depositional sequence. Data based on a suite of sidewall core samples recovered from a Marcellus Formation well; the location is proprietary. MRS = maximum regressive surface; MFS = maximum flooding surface; RST = regressive systems tract; TST = transgressive systems tract.

Overlying MSS1 transgressive systems tract depos- mum flooding surface is roughly coincident with a its display upward-increasing (“dirtying”) gamma- condensed section defined by abundant pyrite and ray and upward-decreasing bulk density log signa- thin carbonate layers, both evident in bulk density tures (Figure 20), principally reflective of decreasing and photoelectric index wireline logs as well as in grain size and increasing TOC (e.g., Singh et al., core. Mineralogic analysis of a suite of sidewall core 2008) related to increasing base level. In the more samples recovered from MSS1 transgressive sys- distal northwestern region of the basin, however, tems tract deposits reveals an abundance of quartz transgressive systems tract deposits appear to be well in excess of that observed in overlying regres- very thin or absent, the contact of the Onondaga sive systems tract deposits (Figure 21). Note that Limestone and overlying Union Springs Member the clay content of the quartz-rich interval is rela- being sharp (see Figure 10D). These relations sug- tively low (Figure 21), likely a reflection of the rapid gest that the rate of base level rise in this region of landward shift of marine environments at this time the basin far exceeded clastic sediment flux. (e.g., Liro et al., 1994). Thin section and scanning The top of the MSS1 transgressive systems electron microscopic examinations reveal that the tract, the maximum flooding surface, is placed at a bulk of the quartz in the MSS1 transgressive sys- gamma-ray peak a short distance above the maxi- tems tract (condensed section) is microcrystalline, mum regressive surface (Figure 20). The maxi- likely derived from the dissolution of silica tests.

84 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 22. Gamma-ray log (left) showing the gradational contact (shaded interval) of the Cherry Valley Member and overlying Oatka Creek Member.

Much of the quartz lines pore throats or forms ir- The MSS1 maximum flooding surface defines regular microcrystalline aggregates that coat de- the top of the transgressive systems tract. Immedi- trital clay grains. Occasional angular detrital quartz ately overlying regressive systems tract deposits re- and feldspar grains are probably windblown detri- cord the slow reduction of base level and/or increased tus. Calcite is as much as three times as abundant sediment flux relative to base level rise (Figure 20). in transgressive systems tract deposits as in the over- Overlying strata display a gradual increase in bulk lying regressive systems tract (Figure 21). Most cal- density and reduced gamma-ray response (Figure 20) cite occurs as single crystals or patches of microspar reflecting reduced organic carbon and increasing and microcrystalline aggregates that originated from clay content. Diminished total quartz (Figure 21) styliolinid fragments. Last, peaks in pyrite and TOC probably records dilution of the biogenic con- are coincident with the inferred maximum flooding tribution by increasing amounts of clastic detritus, surface (Figure 21). At this time, conditions con- principally clay, as nearshore environments were ducive to the preservation of organic matter, per- displaced seaward and accommodation space di- haps fully euxinic bottom conditions related to sa- minished. The reduction in base level reflected in linity and density stratification of the water column the MSS1 regressive systems tract culminated in ac- (e.g., Ettensohn and Elam, 1985; Werne et al., 2002), cumulation of carbonate deposits of the Cherry Val- were established. Alternatively, the relatively rapid ley Member across much of the basin (Figure 11). rate of sedimentation during the time of Marcellus Werne et al. (2002) maintain that bioclastic debris deposition, perhaps enhanced by a basin shallower and detrital carbonate mud that comprise the than generally presumed, may have combined to Cherry Valley was derived, in part, from exposed diminish the rate of oxidation of organic matter in carbonate platform areas bordering the basin to the water column. the west, including the Cincinnati and Algonquin

Lash and Engelder 85 86

Figure 23. Map (A) showing basement structural elements (modified from Alexander et al., 2005) and location of tairpyo h ideDvna aclu omto,AplcinBasin Appalachian Formation, Marcellus Devonian Middle the of Stratigraphy sequence stratigraphic cross-sections (B-G) across the core region of the Mar- cellus Formation basin. RT = Rome trough; L-A = Lawrenceville-Attica cross- structural discontinuity; T-MU = Tyrone- Mount Union cross-structural discontinuity; P-W = Pittsburgh-Washington cross- structural discontinuity. Cross-sections are hung on the top of the MSS2 T-R sequence. Shaded ovals on the map de- note cross-section segments (with shaded labels) that display evidence of syndepositional faulting. Logs are gamma-ray logs; maximum American Petroleum Institute (API) count = 700. Refer to text for discussion. ahadEngelder and Lash

Figure 23. Continued. 87 Figure 23. Continued. arches. Similarly, Brett and Baird (1996) have inter- from the top of the Cherry Valley Member in preted bioclastic carbonates higher in the Hamilton central western New York as a lag deposit formed Group to be lowstand deposits. The maximum re- during maximum transgression, consistent with gressive surface that defines the top of MSS1, then, asymmetric base level curves typified by a rapid is placed at a gamma-ray minimum and/or bulk initial rise in base level (e.g., Embry et al., 2007). density maximum within the Cherry Valley Mem- The wave-worked carbonate deposits are overlain ber (Figure 20). by the increasingly organic-rich (increasing gamma- Note that the Columbia 1 Rodolfy well in ray and deceasing bulk density) basal interval of the Wayne County, Pennsylvania, penetrated ap- Oatka Creek Member. These deposits comprise proximately 65 ft (20 m) of what appears to be the bulk of the MSS2 transgressive systems tract arenaceous limestone at the top of the regressive (Figure 20). systems tract interval (see Figure 4). These depos- A condensed section associated with the MSS2 its, included in the Cherry Valley Member of this maximum flooding surface, which is placed at a study (Figures 3, 4), are interpreted to comprise a gamma-ray maximum and bulk density minimum paralic succession (e.g., Emery and Myers, 1996) (Figure 20), includes carbonate and pyritiferous that had prograded in response to reduced base layers evident on bulk density and photoelectric level. This arenaceous facies of the Cherry Valley index logs. Werne et al. (2002) and Sageman et al. Member in the subsurface of northeastern and (2003), as part of a geochemical investigation of central Pennsylvania and probably northern West Oatka Creek samples recovered from two cores Virginia, as well as in exposure near Kingston, New drilled in western New York, observed a strong York (Ver Straeten and Brett, 2006), approxi- correlation between TOC maxima and intervals of mately 50 mi (80 km) east of the Columbia 1 sediment starvation. However, sidewall core data Rodolfy well, reflects the effects of reduced base of the present study and gamma-ray and bulk den- level at the end of MSS1 time. sity log signatures (see Figure 8)suggestthatMSS2 Well-log signatures suggest a normally rapid transgressive systems tract (condensed section) transition from the Cherry Valley Member into the deposits are not as organic rich as equivalent de- overlying Oatka Creek Member (Figure 22). The posits of the MSS1 T-R sequence. The seemingly upper part of the Cherry Valley represents trans- less organic nature of the MSS2 transgressive sys- gressive carbonate deposits recording the initiation tems tract (condensed section) may reflect the of the MSS2 base level rise. Ver Straeten et al. existence of a better established terrestrial drain- (1994) interpret a prominent bone bed described age network capable of diluting the organic flux by

88 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 23. Continued.

this time. Alternatively, the magnitude of the base tritus to the basin and consequent dilution of or- level rise recorded by MSS2 transgressive systems ganic material raining out of the water column is tract deposits may not have been great enough to reflected in diminishing TOC upsection recorded by shift facies belts far enough landward to effectively a progressive increase in bulk density and reduced isolate the basin from clastic detritus. gamma-ray count (Figure 20). Furthermore, some Regressive systems tract deposits immediately gamma-ray logs display a facies change upward from above the MSS2 maximum flooding surface re- relatively carbonaceous lower regressive systems cord the slowing of base level rise and increasing tract deposits into an overlying organic-lean regres- sediment supply and consequent seaward shift of sive systems tract interval (Figure 20). In places, the shoreline. The increasing supply of clastic de- however, this facies distinction is better made on

Lash and Engelder 89 bulk density logs (Figure 20),theresponseofwhich appears to provide a truer indication of organic carbon content (e.g., Schmoker, 1979, 1980, 1981). The MSS2 base level cycle culminated with accumulation of carbonate deposits of the Stafford Member of the Skaneateles Formation, a likely lowstand carbonate (e.g., Brett and Baird, 1996). In the absence of limestone, the maximum regressive surface delimiting the top of MSS2 is placed at a gamma-ray minimum and/or bulk density maximum (see Figure 16B). Such a contact separating a “cleaning-up trend” from an overlying “dirtying-up trend” has been described from several modern basins, including the Upper Jurassic Ute Field, offshore Norway (Emery and Myers, 1996). A series of sequence-stratigraphic cross sections through the core region of the basin (Figure 23A) reveals several significant aspects of the stratigraphic architecture of the Marcellus Formation. MSS1 is thickest in northeastern Pennsylvania and southeastern New York where a thick regressive systems tract interval includes arenaceous limestone (Figure 23B, C). Similarly, depositional sequence MSS2 thickens toward the northeastern region of the basin, the bulk of the thickening restricted to the regressive systems tract (Figure 23B–D). Moreover, gamma- ray log responses suggest that MSS2 transgressive systems tract deposits are less organic rich in this region of the basin (Figure 23B, C), a likely reflection of the dilution of these deposits by a high clastic flux derived from the Acadian highland source region. This interpretation is buttressed by the increasing thickness of MSS2 transgressive systems tract deposits to the northeast (Figure 23B, C). Impressed upon the northeastward thickening trends of both MSS1 and MSS2 are local variations in thickness that may reflect the effects of basement tectonics (e.g., Harper, 1989). Indeed, the imprint of Rome trough–related extensional tectonics in southwestern Pennsylvania appears to be manifested by variations in the thickness of MSS1 between Washington and Fayette counties (Figure 23F). Furthermore, variations in the thickness of Marcellus Formation

T-R sequences are found in the western region of Continued. Pennsylvania (Figure 23D, E, especially F), proxi- maltotheprojectedregionofRometrough–related faulting (Figure 23A). Figure 23.

90 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin ahadEngelder and Lash

91 Figure 23. Continued. Variations in the thickness of MSS1 regressive systems tract deposits are suggestive of local erosion. Notably, the base of MSS2 cuts progressively deeper into MSS1 westward from eastern New York (Figure 23B, C). Moreover, the absence of MSS1 in some regions of the basin indicates that erosion has occurred. In this case, the base of MSS2 is interpreted to be an unconformable shoreline ravinement that passes laterally into a maximum regressive surface in the conformable succession (Figure 23B–D, G). Elsewhere, the MSS2 maximum flooding surface is no more than a few meters or so above the MSS1 maximum flooding surface (Figure 23B–G), again, indicative of erosion. The thinning and/or local absence of MSS1 in western New York and Pennsylvania may reflect the effects of both local uplift and concomitant lowering of base level. In this scenario, uplift began soon after MSS1 regressive systems tract deposits started to accumulate. This, in tandem with the drop in base level that followed MSS1 maximum flooding, resulted in the accumulation of a thin, perhaps starved, regressive systems tract across the uplifted region of the basin. However, the local absence of Cherry Valley lowstand carbonates (e.g., Figure 8B) indicates that T-R sequence MSS1 was eroded. Regressive systems tract deposits of MSS2, like their counterparts of T-R sequence MSS1, exhibit local variations in thickness (Figure 23D–F). More- over, the upper boundary of MSS2 cuts well down intothesequenceinwesternNewYork(Figure 23B, C, G). Locally, the maximum regressive surface at the top of MSS2 lies within several meters of the maximum flooding surface (Figure 23C, G). How- ever, MSS2 regressive systems tract deposits thicken into northeastern Ohio before thinning to the west (Figure 23C). The marked thinning of MSS2 in western New York, locally evident in the subsurface of western Pennsylvania (Figure 23E), appears to be a consequence of uplift and related sediment starvation across this region of the basin rather than local erosion. The presence of the complete MSS2

Continued. T-R sequence, including lowstand carbonate rocks of the Stafford Member, indicates that erosion is not responsible for the marked thinning of MSS2 in this region of the basin. Instead, thinning is likely Figure 23.

92 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin a consequence of reduced sedimentation rate, per- Ettensohn (1985, 1987, 1994) proposed a tectono- haps related to local uplift. Note that the gamma-ray stratigraphic model for the Acadian orogeny that signature of MSS2 regressive systems tract deposits entails four tectophases. Within this framework, across the area of thinning in western New York is black shale accumulated as a consequence of indicative of higher TOC relative to correlative de- thrust-load–induced subsidence and rapid deep- posits to the east and west (compare Figures 15, 23B). ening of the foreland basin. Still, the conodont- The Stafford and overlying Levanna members of based correlation of some black shale intervals of the the Skaneateles Formation comprise the transgres- Appalachian Basin with depositional sequences in sive systems tract of T-R sequence SKS (Figure 20). Europe, Morocco, and the Cordilleran region of the Rising base level is manifested by an upward- United States suggests a eustatic signature (Johnson increasing gamma-ray response and decreasing bulk et al., 1985; Johnson and Sandberg, 1989). Both density; the SKS maximum flooding surface is placed mechanisms—eustasy and thrust-induced sub- at the gamma-ray peak/bulk density minimum within sidence—likely shaped the Middle and Upper De- the organic-rich Levanna Member (Figure 20). vonian stratigraphic architecture of the core region Note that the SKS maximum flooding surface in of Marcellus exploration in the Appalachian Basin central western New York is only approximately (Werne et al., 2002). However, it is difficult at 20 ft (6 m) above the upper maximum regressive best to assess the relative contributions of each surface of T-R sequence MSS2, whereas farther to driving factor—tectonism and eustasy—separately the west, the SKS maximum flooding surface is (Burton et al., 1987). approximately 60 ft (18 m) above the base of SKS. Lithospheric-flexure models predict that the Similarly, the SKS maximum flooding surface in onset of a foreland thrust-load event and conse- northwestern Pennsylvania is only approximately quent elastic basin subsidence is accompanied by 8 ft (2.5 m) above the top of MSS2. However, in development of a forebulge on the cratonward mar- western New York, the same surfaces are separated gin of the foreland basin (Quinlan and Beaumont, by approximately 35 ft (11 m) (Figure 23G). These 1984; Tankard, 1986a, b; Crampton and Allen, 1995; relationships reflect the onlapping of SKS trans- DeCelles and Giles, 1996). Viscoelastic (Quinlan gressive systems tract deposits to the west and north. and Beaumont, 1984; Beaumont et al., 1987) and elastic flexure (Flemings and Jordan, 1990; Jordan and Flemings, 1991) models envisage both cratonward BASIN DYNAMICS: THE ROLE OF and hinterlandward migration of foredeeps and BASEMENT STRUCTURES ON MARCELLUS bulges that reflect different loading and subsequent FORMATION SEDIMENTATION PATTERNS relaxation histories (e.g., Ettensohn and Chesnut, 1989; Ettensohn, 1992, 1994; Giles and Dickinson, The Middle and Upper Devonian succession of the 1995). However, the Appalachian foreland basin, Appalachian Basin records the cratonward advance at the onset of the Acadian orogeny, hosted a net- of the Catskill Delta complex in response to the work of intermittently active basement structures Acadian oblique collision of the Avalonia micro- inherited from the breakup of Rodinia, the Pre- plate and Laurentia (Ettensohn, 1987). Rapid east- cambrian supercontinent. Chief among these struc- ward thickening of MSS1 and MSS2 (Figure 23C, tures is the Rome trough, which enters Pennsylvania D) toward the Acadian highlands, the remnant of near its southwest corner (Scanlin and Engelder, which is found in New England, is characteristic of 2003), although its location throughout much of foreland basin deposits (DeCelles and Giles, 1996). the state remains elusive (Harper, 1989). Further- Superimposed on this protracted shallowing trend more, the core region of Marcellus exploration in- is a hierarchy of black shale–based depositional se- cludes several northwest-striking basement wrench quences that reflect shorter term base level oscil- faults (Figure 23A) (Parrish and Lavin, 1982; Rodgers lations (e.g., Johnson et al., 1985; Brett and Baird, and Anderson, 1984; Harper, 1989). These faults, 1986, 1996; Van Tassell, 1994; Ver Straeten, 2007). referred to as cross-strike structural discontinuities

Lash and Engelder 93 Figure 24. Maps showing cross-structural discontinuities and (A) the location of the Oriskany no-sand area, Kane gravity high, and the area of the basin absent in the Union Springs Member of the Marcellus Formation; (B) the area of the basin absent in the Cherry Valley Member of the Marcellus Formation; (C) the region of the basin over which the Oatka Creek Member of the Marcellus Formation thins; (D) the distribution of the Stafford Member of the Skaneateles Forma- tion; (E) the distribution of the Levanna Member of the Skaneateles Formation illustrating 0 ft, 5 ft (1.5 m), and 20 ft (6.1 m) isochores. Cross- structural discontinuities: L-A = Lawrenceville- Attica; B-B = Blairsville-Broadtop; H-G = Home- Gallitzin; T-MU = Tyrone-Mount Union; P-W = Pittsburgh-Washington (modified from Parrish and Lavin, 1982).

94 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Figure 24. Continued.

Lash and Engelder 95 Figure 24. Continued.

([CSD] Wheeler, 1980), likely originated as strike- (Fettke, 1938). Williams and Bragonier (1974), slip faults related to the Late Precambrian forma- commenting on the coincidence of the so-called tion of the proto–Atlantic Ocean (Thomas, 1977). “Oriskany no-sand area” with the Kane gravity high Some authors relate the apparent segmentation (Parrish and Lavin, 1982) (Figure 24A), specu- of the Rome trough to slip on CSD (Lavin et al., lated that the local absence of the Oriskany was a 1982; Harper, 1989); however, it appears that consequence of Early Devonian basement uplift. strike-slip displacement was replaced by dip-slip dis- Ettensohn (1985) subsequently attributed the late placement by early Paleozoic time (Wagner, 1976). Emsian (late Early Devonian) unconformity at the Although the regional architecture of the Acadian base of the Onondaga Limestone to basinward mi- foreland basin was a consequence of load-induced gration of a forebulge caused by tectophase I thrust subsidence, it is difficult to imagine that inherited loading. Similarly, Ver Straeten and Brett (2000) basement structures, including those previously de- related northeastward time transgressive pinnacle scribed, did not, in some way, affect foreland basin reef development described from the upper part of evolution and sedimentation patterns during the the Onondaga Limestone to retrogradational (east- Acadian orogeny. Reactivation of preexisting faults ward) migration of a flexural welt at the end of during foreland flexure can partition the basin into tectophase I of the Acadian orogeny. However, regions of fault-controlled uplift and depocenters the subparallel orientation of the Oriskany no-sand (e.g., Tankard, 1986a; DeCelles and Giles, 1996). area, which is roughly mimicked by Onondaga pin- An early indicator of forebulge-like dynamics nacle reef formation (Figure 24A), and the oblique induced by Acadian convergence in the Appala- Acadian plate convergence direction (Ettensohn, chian Basin is the thinning and local absence of the 1987; Ferrill and Thomas, 1988) is inconsistent late Early Devonian Oriskany Sandstone along a with erosion of the Oriskany as a straightforward northeast-southwest–trending region of north cen- flexural response to thrust loading in the hinter- tral and northwestern Pennsylvania (Figure 24A) land (i.e., Turcotte and Schubert, 1982). Ver Straeten

96 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin and Brett (2000) suggested that the inferred pos- the basin, however, farther removed from clastic itive feature on which Onondaga pinnacle reefs sources, rising base level was not accompanied by formed does not comport with regional forebulge deposition of an obvious transgressive systems tract. models. Instead, the rise in base level is manifested only by Note that the elongate Oriskany no-sand area a sharp contact of the Onondaga Formation and and Kane gravity high are roughly centered be- overlying Union Springs Member (see Figure 10D). tween the Lawrenceville-Attica and Home-Gal- Reactivated basement structures, including Rome litzin CSD (Figure 24A). Furthermore, Onondaga trough–like extensional structures, appear to have pinnacle reef development appears to have termi- influenced the thickness of MSS1 and MSS2 from nated close to the trace of the Lawrenceville-Attica southwestern into central Pennsylvania (Figure 23D– CSD (Figure 24A). It is an intriguing possibility that F). The Union Springs Member, which comprises thickness trends of the Lower Devonian Oriskany the bulk of MSS1 transgressive and regressive sys- Sandstone and Onondaga reef development in this tems tract deposits, as well as the Cherry Valley region of the basin reflect episodes of vertical dis- Member, much of which encompasses the upper placement of crustal blocks bounded by the Home- part of the MSS1 regressive systems tract, are ab- Gallitzin, Tyrone-Mount Union, and Lawrenceville- sent along a northeast-southwest–oriented region Attica CSD (Figure 24A). Such block displacement, of western New York and northwest Pennsylvania which would have overprinted smooth elastic fore- (Figures 5, 11, 24B) that parallels the Oriskany no- bulge dynamics, partitioned the foredeep basin into sand area to the south (Figure 23A, B). As with the subtle ridges and depocenters (DeCelles and Giles, Oriskany no-sand area, the western and eastern limits 1996). We have no data to suggest that the faults of the region absent in the Union Springs and were anything but blind (i.e., did not intersect the Cherry Valley (MSS1) deposits are roughly coin- Earth’s surface) during the time of Marcellus depo- cident with the Home-Gallitzin and Lawrenceville- sitions. Still, enough displacement may have oc- Attica CSD, respectively (Figure 24A, B). The thin- curred such that related base level changes produced ning and local absence of MSS1 deposits likely local sites of wave base sediment reworking (i.e., reflects the effects of local warping or flexing of the unconformable shoreline ravinement) adjacent to basin induced by displacement on the Lawrenceville- depocenters. Some authors have suggested that the Attica and Home-Gallitzin CSD soon after the Marcellus Formation accumulated in deep oxygen- MSS1 base level maximum. The result was a local deficient waters (Potter et al., 1981), yet a rela- reduction of base level of such a magnitude to cause tively shallow water depth, perhaps only 100 to erosion of MSS1. 130 ft (30–40 m), would have reduced the vertical Transgressive systems tract deposits of MSS2, displacement necessary to bring base level to such the upper reworked horizon of the Cherry Valley a position that local sediment starvation or even Member and radioactive basal interval of the over- erosioncouldhaveoccurredwithinthelimitedtime lying Oatka Creek Member, record a rapid base framework over which the Marcellus accumulated. level rise. Overlying regressive systems tract strata Indeed, Harper (1999) has suggested that the sub- (upper Oatka Creek Member and lower interval of tropical Marcellus Basin could have been less than the Stafford Member of the Skaneateles Formation) 165 ft (50 m) deep if the water column had been thicken markedly toward the northeastern region of sufficiently stratified to preclude mixing of warm the basin (Figure 23B, C), a likely consequence of oxygenated surface water with anoxic or, as sug- increasing proximity to the thrust load and clastic gested by Werne et al. (2002), euxinic bottom water. sources (e.g., DeCelles and Giles, 1996). Accom- The organic-rich Union Springs Member re- modationspaceatthistimewouldhavebeencre- flects a sharp increase in base level that resulted ated by relaxation of the tectonic load (Ettensohn, in accumulation of a transgressive systems tract at 1994). However, like MSS1, the MSS2 sequence the onset of tectophase II of the Acadian orogeny thins along a northeast-southwest–trending region (Ettensohn, 1985, 1994). In more distal regions of of western New York and Pennsylvania (Figures 12,

Lash and Engelder 97 23B). Note also that the region of thin MSS2 de- tion, records the base level minimum at the top of posits, notably the Oatka Creek Member, parallels the MSS2 T-R sequence. The isochore pattern of the Oriskany no-sand area and is roughly bounded the Stafford, too, may reflect the influence of fault- on the west by the Tyrone-Mount Union CSD induced warping of the basin (Figure 17). That is, (Figure 24C). The northern and eastern limit of the the southwestern edge of the Stafford Member ze- region of Oatka Creek thinning appears to have roes close to the inferred trace of the Tyrone-Mount been lost to erosion (Figure 24C). The presence of Union CSD; to the east, the Lawrenceville-Attica the complete MSS2 T-R sequence in the areas of CSD appears to have exerted little, if any, control thinning in western New York (Figure 23B, C, G) on accumulation of the carbonate lowstand de- indicates that thinning was a consequence of re- posits (Figure 24D). However, the Stafford is duced sediment flux rather than erosion. Moreover, thickest in the center of the block bounded by the the thin MSS2 sequence, especially the regressive Tyrone-Mount Union and Lawrenceville-Attica CSD systems tract deposits, is more organic rich than (Figures 17, 24D) perhaps revealing the shallowest laterally adjacent thicker MSS2 deposits (compare region of the block at this time. Last, the south- Figures 15, 23B). We attribute local thinning of western extent of the Stafford may reflect the influ- MSS2 to accumulation of a starved or condensed T-R ence of the Blairsville–Broadtop CSD (Figure 24D). sequence across a northeast-southwest–trending to- Conceivably, a complex displacement history on pographic high. In this scenario, organic-rich se- both the Tyrone-Mount Union and Blairsville- diment raining out of the water column was con- Broadtop CSD influenced sedimentation patterns centrated on the high even as base level dropped, at the western edge of the Stafford depocenter. whereas organic-lean sediment, perhaps deposited The base level rise after accumulation of MSS2 from hyperpycnal flows, ponded in adjacent bathy- regressive systems tract deposits is reflected in the metric lows. Levanna Member of the Skaneateles Formation, The story that emerges from the described transgressive systems tract deposits of the SKS T-R thickness trends is one of local basement control on sequence (Figure 20). The especially intriguing Marcellus sedimentation patterns that began at the aspect of the Levanna isochore pattern is its sharp end of Early Devonian time in the northwestern eastern termination close to the inferred trace of the Pennsylvania region of the basin with erosion of the Lawrenceville-Attica CSD (Figure 24E). Relatively Oriskany Sandstone. Syndepositional movement rapid thinning of the organic-rich Levanna to the on what were likely blind faults resulted in the east (Figure 18) is suggestive of down-to-west dis- creation of local depocenters and subtle ridges that placement along the Lawrenceville-Attica fault, influenced sedimentation and erosion patterns in creating a black shale depocenter. The western limit this region of the greater Acadian foreland basin of the Levanna is more gradual (Figures 18, 24E). that had formed as a consequence of thrust load- The thicker region of the Levanna (20-ft [6.1 m] ing. The block delimited by the Home-Gallitzin isochore) terminates close to the Tyrone-Mount and Lawrenceville-Attica CSD experienced epi- Union CSD (Figure 24E). However, the effects of sodes of uplift, resulting in localized erosion and/or this fault on basin morphology appear not to have sediment starvation across a topographic high fol- been great enough to preclude accumulation of lowed by subsidence, the creation of accommoda- carbonaceous sediment well to the southwest of tion space, and consequent sediment accumula- the Tyrone-Mount Union CSD. The southwestern tion. Block activation appears to have shifted to the limit of Levanna sedimentation may have been northwest and then to the east during the time of influenced by down-to-east displacement on the Marcellus deposition. Furthermore, the history of Pittsburgh-Washington CSD (Figure 24E). Pro- crustal block movement, apparently linked to ob- gressive subsidence of the crustal block defined lique plate convergence, continued beyond accumu- by the Lawrenceville-Attica and Tyrone-Mount lation of the Marcellus Formation. The Stafford Union CSD in tandem with rising base level is in- Member, the basal unit of the Skaneateles Forma- dicated by westward and northward onlapping of

98 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin the most organic-rich facies of the Levanna Mem- ness, including quartz, calcite, and pyrite. Assessing ber (Figure 23G). therelativeimportanceofthesecompositionalele- The generally limited distribution of the organic- ments to production and stimulation strategies be- rich Levanna Member relative to the more wide- comes a reservoir engineering issue. spread Union Springs and Oatka Creek members Variations in the thickness of lithostratigraphic of the Marcellus Formation appears to have been a units of the Marcellus Formation and immediately consequence of reactivated blind basement struc- overlying deposits of the Skaneateles Formation as tures and consequent warping of the foreland basin. well as the MSS1 and MSS2 T-R sequences across Specifically, displacement along the Lawrenceville- the core region of the basin reflect the complex Attica, Tyrone-Mount Union, and Pittsburgh- relationship among Acadian thrust loading, fluc- Washington basement wrench faults created the ac- tuations in base level, recurrent basement struc- commodation space necessary for accumulation of tures, and proximity to clastic sources. Acadian carbonaceous sediment in a relatively restricted re- thrust loading of Laurentia played a first-order role gion of the greater Acadian foreland basin, which, at in creating the accommodation space necessary for this time, was receiving otherwise organic-lean sedi- accumulation of both Marcellus T-R sequences. ment of the Skaneateles Formation and equivalents. Indeed, accommodation space was greatest in the northeast region of the basin, proximal to the Aca- dian thrust load and clastic sources. However, CONCLUSIONS marked local variations inthethicknessofboth T-R sequences, especially regressive systems tract The stratigraphy and thickness trends of the Mid- deposits, are likely a consequence of displacement dle Devonian Marcellus Formation of the Appala- along reactivated blind basement faults, includ- chian Basin reflect the interplay of Acadian thrust ing those associated with the Rome trough, which loading of the Laurentian craton and short-term warped the foreland basin into low relief ridges and base level fluctuations. These events shaped the depocenters. stratigraphic architecture of the Marcellus in ways The possible influence of structural control on that can impact exploration and production strat- the thickness of the Marcellus Formation is not a egies of this emerging shale gas play. Analysis of new concept. Piotrowski and Harper (1979) pre- more than 900 wireline logs indicates that the sented evidence that the Laurel Hill, Negro Moun- Marcellus Formation encompasses two third- tain, and Chestnut Ridge anticlines of southwest order T-R sequences, MSS1 and MSS2, in ascending Pennsylvania were active during accumulation of order. These deposits are overlain by T-R sequence the Marcellus Formation. More recently, Scanlin SKS, which comprises at least the lower part of the and Engelder (2003) demonstrated that the thick- Skaneateles Formation. The Marcellus sequence ening of Marcellus black shale in the hinges of anti- stratigraphy offers a predictive framework for res- clines in southwest Pennsylvania was a consequence ervoir assessment that can be extrapolated into areas of Alleghanian folding. The present study extends of poor data control. Compositional attributes that the concept of basement control on sedimentation influence such critical reservoir properties as po- and erosion patterns over a greater expanse of the rosity and brittleness, including quartz, carbonate, basin. The northwest-striking TyroneMount Union, clay, and pyrite, vary predictably as a consequence Lawrenceville-Attica, and Home-Gallitzin wrench of base level oscillations. The sequence-stratigraphic faults, most likely blind, appear to have been es- framework of the Marcellus Formation presented pecially active in the central to western New York in this study demonstrates that transgressive systems and western Pennsylvania region of the Appala- tract and early regressive systems tract deposits con- chian Basin during the time of Marcellus deposition, tain the greatest abundance of malleable organic producing subtle ridges and depocenters manifested matter. However, these same deposits are enriched by local erosion and increases in thickness of the in those components that enhance reservoir brittle- MSS1 and MSS2 sequences.

Lash and Engelder 99 REFERENCES CITED Creaney, S., and Q. R. Passey, 1993, Recurring patterns of total organic carbon and source rock quality within a se- Alexander, S. S., R. Cakir, A. G. Doden, D. P. Gold, and S. I. quence stratigraphic framework: AAPG Bulletin, v. 77, Root,compilers,2005,Basementdepthandrelated p. 386–401. geospatial database for Pennsylvania: Pennsylvania Geo- Curiale, J. A., R. D. Cole, and R. J. Witmer, 1991, Applica- logical Survey, 4th series, Open-File General Geology tion of organic geochemistry to sequence stratigraphic Report 05-01.0, www.dcnr.state.pa.us/topogeo/openfile analysis: Four Corners area, New Mexico, U.S.A.: Organic (accessed January 29, 2010). Geochemistry, v. 19, p. 53–75, doi:10.1016/0146-6380 Beaumont, C., G. M. Quinlan, and J. Hamilton, 1987, The (92)90027-U. Alleghanian orogeny and its relationship to the evolution DeCelles, P. G., and K. A. Giles, 1996, Foreland basins: Basin of the eastern interior, North America, in C. Beaumont Research,v.8,p.105–123, doi:10.1046/j.1365-2117 and A. J. Tankard, eds., Sedimentary basins and basin- .1996.01491.x. forming mechanisms: Canadian Society of Petroleum Dennison, J. M., and K. O. Hasson, 1976, Stratigraphic cross Geologists, Memoir 12, p. 425–446. section of Hamilton Group (Devonian) and adjacent strata Brett, C. E., and G. C. Baird, 1986, Symmetrical and upward along south border of Pennsylvania: AAPG Bulletin, v. 60, shoaling cycles in the Middle Devonian of New York p. 278–287. State and their implications for the punctuated aggrada- de Witt Jr., W., W. J. Perry Jr., and L. G. Wallace, 1975, Oil tional cycle hypothesis: Paleooceanography, v. 1, p. 431– and gas data from Devonian and Silurian rocks in the Ap- 445, doi:10.1029/PA001i004p00431. palachian Basin: U.S. Geological Survey Miscellaneous Brett, C. E., and G. C. Baird, 1996, Middle Devonian sedi- Investigations Series Map I-917 B. mentary cycles and sequences in the northern Appala- de Witt Jr., W., J. B. Roen, and L. G. Wallace, 1993, Stratigra- chian Basin, in B.J.Witzke,G.A.Ludvigson,andJ. phy of Devonian black shales and associated rocks in the Day, eds., Paleozoic sequence stratigraphy: Views from Appalachian Basin: U.S. Geological Survey Bulletin 1909, the North American craton: Geological Society of Amer- p. B1–B47. ica Special Paper 306, p. 213–241. Embry, A., 1993, Trangressive-regressive (T-R) sequence Brett, C. E., and C. A. Ver Straeten, 1995, Middle Devonian analysis of the Jurassic succession of the Sverdrup Basin, (Eifelian) carbonates, northern and central Appalachian Canadian Arctic Archipelago: Canadian Journal of Earth Basin: Sequence stratigraphic framework: AAPG Bulletin, Science, v. 30, p. 301–320. v. 79, p. 1410–1411. Embry, A., 1995, Sequence boundaries and sequence hierar- Brown Jr., L. F., R. G. Loucks, and R. H. Trevino, 2005, Site- chies: Problems and proposals, in R. J. Steel, V. L. Felt, E. specific sequence-stratigraphic section benchmark charts P. Johannesson, and C. Mathieu, eds., Sequence stratigra- are key to regional chronostratigraphic systems tract anal- phy on the northwest European margin: Norwegian Pe- ysis in growth-fault basins: AAPG Bulletin, v. 89, p. 715– troleum Society Special Publication 5, p. 1–11. 724, doi:10.1306/01270504066. Embry, A., 2002, Transgressive-regressive (T-R) sequence Burton, R., C. Kendall, and I. Lerche, 1987, Out of our depth: stratigraphy, in J. Armentrout and N. Rosen, eds., Gulf On the impossibility of fathoming eustasy from the strati- Coast SEPM Conference Proceedings, Houston, Texas, graphic record: Earth Science Reviews, v. 24, p. 237– p. 151–172. 277, doi:10.1016/0012-8252(87)90062-6. Embry, A., and E. Johannessen, 1992, T-R sequence stratigra- Cate, A. S., 1963, Lithostratigraphy of some Middle and phy, facies analysis and reservoir distribution in the upper- Upper Devonian rocks in the subsurface of southwestern most Triassic-Lower Jurassic succession, western Sverdrup Pennsylvania: Pennsylvania Topographic and Geological Basin, Arctic Canada, in T. O. Vorren, E. Bergsager, Ø. A. Survey General Geology Report, Fourth Series, No. G39, Dahl-Stamnes, E. Holter, B. Johansen, E. Lie, and T. B. p. 229–440. Lund, eds., Arctic geology and petroleum potential: Nor- Clarke, J. M., 1903, Classification of New York series of geo- wegian Petroleum Society Special Publication 2, p. 121– logic formations: New York State Museum Handbook 146. No. 19, table 2, 28 p. Embry, A., E. Johannessen, D. Owen, B. Beauchamp, and P. Cooper, G. A., 1930, Stratigraphy of the Hamilton Group Gianolla, 2007, Sequence stratigraphy as a “concrete” of New York: American Journal of Science, 5th series, stratigraphic discipline: Report of the ISSC Task Group v. 19, p. 116–134, 214-236. on Sequence Stratigraphy, 104 p., http://sepmstrata.org Cooper, G. A., 1933, Stratigraphy of the Hamilton Group of /PDF-Files/SeqStrat-Refs/ISSC_Report_on_Sequence eastern New York, part 1: American Journal of Science, _Stratigraphy.pdf (accessed February 16, 2010). v. 26, p. 537–551. Emery, D., and K. J. Myers, 1996, Sequence stratigraphy: Cooper, G. A., 1934, Stratigraphy of the Hamilton Group of Blackwell Science, Oxford, United Kingdom, 297 p. eastern New York, part 2: American Journal of Science, Engelder, T., G. G. Lash, and R. S. Uzcátegui, 2009, Joint sets v. 27, p. 1–12. that enhance production from Middle and Upper Devo- Crampton, S. L., and P. A. Allen, 1995, Recognition of fore- nian gas shales of the Appalachian Basin: AAPG Bulletin, bulge unconformities associated with early stage foreland v. 93, p. 857–889, doi:10.1306/03230908032. basin development: Example from the North Alpine Ettensohn, F. R., 1985, The Catskill delta complex and the foreland basin: AAPG Bulletin, v. 79, p. 1495–1514. Acadian orogeny: A model, in D. L. Woodrow and W. D.

100 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Sevon, eds., The Catskill delta: Geological Society of Harper, J. A., and R. G. Piotrowski, 1979, Stratigraphic cor- America Special Paper 201, p. 39–49. relation of surface and subsurface Middle and Upper De- Ettensohn, F. R., 1987, Rates of relative plate motion during vonian, southwestern Pennsylvania, in J. M. Dennison, the Acadian orogeny based on the spatial distribution of K. O. Hasson, D. M. Hoskins, R. M. Jolley, and W. D. blackshales:JournalofGeology,v.95,p.572–582, Sevon, eds., Devonian shales in south-central Pennsylva- doi:10.1086/629150. nia and Maryland: 44th Annual Field Conference of Ettensohn, F. R., 1992, Controls on the origin of the Devonian– Pennsylvania Geologists Guidebook, Bedford, Pennsyl- Mississippian oil and gas shales, east-central United States: vania, p. 18–37. Fuel, v. 71, p. 1487–1492, doi:10.1016/0016-2361(92) Jervey, M. T., 1988, Quantitative geological modeling of 90223-B. siliciclastic rock sequences and their seismic expression, Ettensohn, F. R., 1994, Tectonic controls on formation and in C. K. Wilgus, B. S. Hastings, C. G. St. C. Kendall, H. cyclicity of major Appalachian unconformities and asso- Posamentier, C. A. Ross, and J. Van Wagoner, eds., Sea- ciated stratigraphic sequences: SEPM Concepts in Sedi- level changes: An integrated approach: SEPM Special mentology and Paleontology No. 4, Tectonic and Eu- Publication 42, p. 47–69. static Controls on Sedimentary Cycles, p. 217–242. Johnson, J. G., and C. A. Sandberg, 1989, Devonian eustatic Ettensohn, F. R., and D. R. Chestnut Jr., 1989, Nature and events in the western United States and their biostrati- probable origin of the Mississippian–Pennsylvania un- graphic responses, in N. J. McMillan, A. F. Embry, and conformity in the eastern United States: Internationale D. J. Glass, eds., Devonian of the world: Canadian Soci- Congres de Stratigraphic et Geologie du Carbonifere, ety of Petroleum Geologists, Memoir 14, v. 3, p. 171– Beijing, 11th, Compte Rendu 4, p. 145–159. 178. Ettensohn, F. R., and T. D. Elam, 1985, Defining the nature Johnson, J. G., G. Klapper, and C. A. Sandberg, 1985, Devo- and location of a Late Devonian–Early Mississippian pycno- nian eustatic fluctuations in Euramerica: Geological So- cline in eastern Kentucky: Geological Society of America ciety of America Bulletin, v. 96, p. 567–587, doi:10 Bulletin, v. 96, p. 1313–1321, doi:10.1130/0016-7606 .1130/0016-7606(1985)96<567:DEFIE>2.0.CO;2. (1985)96<1313:DTNALO>2.0.CO;2. Jordan, T. E., and P. B. Flemings, 1991, Large scale strati- Faill, R. T., 1985, The Acadian orogeny and the Catskill Delta, in graphic architecture, eustatic variation, and unsteady tec- D. L. Woodrow and W. D. Sevon, eds., The Catskill Delta: tonism: A theoretical evaluation: Journal of Geophysical Geological Society of America Special Paper 201, p. 15–38. Research, v. 96, p. 6681–6699, doi:10.1029/90JB01399. Ferrill, B. A., and W. A. Thomas, 1988, Acadian dextral trans- Kamakaris, D. G., and A. M. Van Tyne, 1980, Isopach map of pression and synorogenic sedimentary successions in the black shale in the Hamilton Group (from well sample stud- Appalachians: Geology, v. 16, p. 604–608, doi:10.1130 ies): Morgantown Energy Technology Center, Morgan- /0091-7613(1988)016<0604:ADTASS>2.3.CO;2. town Energy Tech Control/Eastern Shale Gas Project Fettke, C. R., 1938, Oriskany as a source of gas and oil in (METC/ESGP) Series 110. Pennsylvania and adjacent areas: AAPG Bulletin, v. 22, Kaufmann, B., 2006, Calibrating the Devonian time scale: A p. 241–266. synthesis of U-Pb ID-TIMS ages and conodont stratigra- Flemings, P. B., and T. E. Jordan, 1990, A synthetic strati- phy: Earth Science Reviews, v. 75, p. 175–190, doi:10 graphic model of foreland basin development: Journal of .1016/j.earscirev.2006.01.001. Geophysical Research, v. 94, p. 3851–3866, doi:10.1029 Lambert, M. W., 1993, Internal stratigraphy and organic fa- /JB094iB04p03851. cies of the Devonian-Mississippian Chattanooga (Wood- Giles, K. A., and W. R. Dickinson, 1995, The interplay of ford) shale in Oklahoma and Kansas, in B. J. Katz and eustasy and lithospheric flexure in forming stratigraphic L. M. Pratt, eds., Source rocks in a sequence stratigraphic sequences in foreland settings: An example from the framework: AAPG Studies in Geology 37, p. 163–176. Antler foreland, Nevada and Utah, in S. L. Dorobek Lavin, P. M., D. L. Chaffin, and W. Davis, 1982, Major linea- and G. M. Ross, eds., Stratigraphic evolution of foreland ments and the Lake Erie-Maryland crustal block: Tec- basins: SEPM Special Publication 52, p. 187–209. tonics, v. 1, p. 431–440, doi:10.1029/TC001i005p00431. Hall, J., 1839, Third annual report of the Fourth Geological Leckie, D. A., C. Singh, F. Goodarzi, and J. H. Wall, 1990, District of the State of New York: Geological Survey of Organic-rich, radioactive marine shale: A case study of a New York Annual Report, v. 3, p. 287–339. shallow-water condensed section, Cretaceous Shaftsbury Harper, J. A., 1989, Effects of recurrent tectonic patterns on Formation, Alberta, Canada: Journal of Sedimentary Pe- the occurrence and development of oil and gas resources trology, v. 60, p. 101–117. in western Pennsylvania: Northeastern Geology, v. 11, Liro, L. M., W. C. Dawson, B. J. Katz, and V. D. Robison, p. 225–245. 1994, Sequence stratigraphic elements and geochemical Harper, J. A., 1999, Devonian, in G. H. Shultz, ed., The geol- variability within a “condensed section”: Eagle Ford ogy of Pennsylvania: Pennsylvania Geological Survey Group, east-central Texas: Transactions of the Gulf Coast Special Publication 1, p. 109–127. Association of Geological Societies, v. 44, p. 393–402. Harper, J. A., and R. G. Piotrowski, 1978, Stratigraphy, ex- Loutit, T. S., J. Hardenbol, P. R. Vail, and G. R. Baum, 1988, tent, gas production, and future gas potential of the De- Condensed sections: The key to age determination and cor- vonian organic-rich shales in Pennsylvania: Preprints, 2d relation of continental margin sequences, in C. K. Wilgus, Eastern Gas Shales Symposium, METC/SP-78/6, v. I, B. S. Hastings, C. G. St. C. Kendall, H. W. Posamentier, p. 310–329. C. A. Ross, and J. C. Van Wagoner, eds., Sea-level

Lash and Engelder 101 changes: An integrated approach: SEPM Special Publica- H. W. Posamentier, C. A. Ross, and J. C. Van Wagoner, tion 42, p. 183–213. eds., Sea-level changes: An integrated approach: SEPM Mancini, E. A., and T. M. Puckett, 2002, Transgressive- Special Publication 42, p. 110–124. regressive cycles in lower Cretaceous strata, Mississippi Potter, P. E., J. B. Maynard, and W. A. Pryor, 1981, Sedimen- Interior Salt Basin area of the northeastern Gulf of Mex- tology of gas-bearing Devonian shales of the Appalachian ico, U.S.A.: Cretaceous Research, v. 23, p. 409–438, Basin: U.S. Department of Energy, USDOE/METC-114, doi:10.1006/cres.2002.1012. Morgantown, West Virginia, 43 p. Mancini, E. A., B. W. Tew, and R. M. Mink, 1993, Petroleum Potter, P. E., J. B. Maynard, and W. A. Pryor, 1982, Appala- source rock potential of Mesozoic condensed section de- chian gas bearing Devonian shales: Statements and dis- posits of southwest Alabama, in B. J. Katz and L. M. cussions:OilandGasJournal,v.80(January25),p.290– Pratt, eds., Source rocks in a sequence stratigraphic frame- 318. work: AAPG Studies in Geology 37, p. 147–162. Quinlan, G. M., and C. Beaumont, 1984, Appalachian thrust- Millbrooke, A., 1981, Henry Darwin Rogers and the first state ing, lithospheric flexure, and the Paleozoic stratigraphy of geological survey of Pennsylvania: Northeastern Geol- the eastern interior of North America: Canadian Journal of ogy, v. 3, p. 71–74. Earth Sciences, v. 31, p. 973–996, doi:10.1139/e84-103. Mitchum, R. M. J., and J. C. Van Wagoner, 1991, High- Rast, N., and J. W. Skehan, 1993, Mid-Paleozoic orogenesis frequency sequences and their stacking patterns: Se- in the North Atlantic: The Acadian orogeny, in D. C. quence stratigraphic evidence of high-frequency eustatic Roy and S. J. Skehan, eds., The Acadian orogeny: Geo- cycles: Sedimentary Geology, v. 70, p. 131–160, doi:10 logical Society of America Special Paper 275, p. 1–25. .1016/0037-0738(91)90139-5. Rickard, L. V., 1984, Correlation of the subsurface Lower and Oliver Jr., W. A., 1954, Stratigraphy of the Onondaga Lime- Middle Devonian of the Lake Erie region: Geological So- stone (Devonian) in central New York: Geological Soci- ciety of America Bulletin, v. 95, p. 814–828, doi:10 ety of America Bulletin, v. 65, p. 621–652, doi:10.1130 .1130/0016-7606(1984)95<814:COTSLA>2.0.CO;2. /0016-7606(1954)65[621:SOTOLD]2.0.CO;2. Rickard, L. V., 1989, Stratification of the subsurface low and Oliver Jr., W. A., 1956, Stratigraphy of the Onondaga Lime- mid-Devonian of New York, Pennsylvania, Ohio, and stone in eastern New York: Geological Society of Amer- Ontario: New York State Museum and Science Map ica, v. 67, p. 1441–1474, doi:10.1130/0016-7606(1956) and Chart Series 39, 59 p. 67[1441:SOTOLI]2.0.CO;2. Rodgers, M. R., and T. H. Anderson, 1984, Tyrone-Mt. Union OliverJr.,W.A.,W.deWittJr.,J.M.Dennison,D.M.Hoskins, cross-strike lineament of Pennsylvania: A major Paleozoic and J. W. Huddle, 1969, Correlation of Devonian rock basement fracture and uplift boundary: AAPG Bulletin, units in the Appalachian Basin: U.S. Geological Survey, v. 68, p. 92–105. Oil and Gas Investigation Chart OC-64. Roen, J. B., L. G. Wallace, and J. de Witt Jr., 1978, Preliminary Palsey, M. A., W. A. Gregory, and G. F. Hart, 1991, Organic stratigraphic cross section sowing radioactive zones in the matter variations in transgressive and regressive shales: Devonian black shales in the central part of the Appalachian Organic Geochemistry, v. 17, p. 483–501, doi:10.1016 Basin: U.S. Geological Survey, Oil and Gas Investigations /0146-6380(91)90114-Y. Chart OC-87. Parrish, J. B., and P. M. Lavin, 1982, Tectonic model for kim- Sageman, B. B., A. E. Murphy, J. P. Werne, C. A. Ver Straeten, berlite emplacement in the Appalachian Plateau of Penn- D. J. Hollander, and T. W. Lyons, 2003, A tale of shales: sylvania: Geology, v. 10, p. 344–347, doi:10.1130/0091 The relative roles of production, decomposition, and dilu- -7613(1982)10<344:TMFKEI>2.0.CO;2. tion in the accumulation of organic-rich strata, Middle- Partington, M. A., B. C. Mitchener, N. J. Milton, and A. J. Upper Devonian, Appalachian Basin: Chemical Geology, Fraser, 1993, Genetic sequence stratigraphy for the v. 195, p. 229–273, doi:10.1016/S0009-2541(02)00397-2. North Sea Late Jurassic and Early Cretaceous: Distribu- Sarg, J. F., 1988, Carbonate sequence stratigraphy, in C. K. tion and prediction of Kimmeridgian—late Ryazanian re- Wilgus, B. S. Hastings, C. G. St. C. Kendall, H. Posamentier, servoirs in the North Sea and adjacent areas, in J. R. Park- C. A. Ross, and J. Van Wagoner, eds., Sea-level changes: er, ed., Petroleum geology of northwest Europe: An integrated approach: SEPM Special Publication 42, Proceedings of the 4th Conference: Geological Society p. 155–181. (London), p. 347–370. Scanlin, M. A., and T. Engelder, 2003, The basement versus Passey, Q. R., S. Creaney, J. B. Kulla, F. J. Moretti, and J. D. the no-basement hypotheses for folding within the Ap- Stroud,1990,Apracticalmodelfororganicrichness palachian Plateau detachment sheet: American Journal from porosity and resistivity logs: AAPG Bulletin, v. 74, of Science, v. 303, p. 519–563, doi:10.2475/ajs.303.6.519. p. 1777–1794. Schmoker, J. W., 1979, Determination of organic content of Piotrowski,R.G.,andJ.A.Harper,1979,Blackshaleand Appalachian Devonian shales from formation-density sandstone facies of the Devonian “Catskill” clastic wedge logs: AAPG Bulletin, v. 63, p. 1504–1509. in the subsurface of western Pennsylvania: U.S. Department Schmoker, J. W., 1980, Organic content of Devonian shale of Energy, Eastern Gas Shales Project, EGSP Series 13, in western Appalachian Basin: AAPG Bulletin, v. 64, 40 p. p. 2156–2165. Posamentier, H. W., M. T. Jervey, and P. R. Vail, P. R., 1988, Schmoker, J. W., 1981, Determination of organic-matter con- Eustatic controls on clastic deposition. I. Conceptual frame- tent of Appalachian Devonian shales from gamma-ray work, in C.K.Wilgus,B.S.Hastings,C.G.St.C.Kendall, logs: AAPG Bulletin, v. 65, p. 1285–1298.

102 Stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin Singh, P., R. Slatt, and W. Coffey, 2008, Barnett Shale— resolution correlation of time and facies: AAPG Methods unfolded: Sedimentology, sequence stratigraphy, and re- in Exploration Series 7, 55 p. gional mapping: Gulf Coast Association of Geological Ver Straeten, C. A., 2007, Basinwide stratigraphic synthesis Societies Transactions, v. 58, p. 777–795. and sequence stratigraphy, upper Pragian, Emsian and Tankard, A. J., 1986a, On the depositional response to Eifelian stages (Lower to Middle Devonian), Appala- thrusting and lithospheric flexure: Examples from the chian Basin, in R. T. Becker and W. T. Kirchgasser, eds., Appalachian and Rocky Mountain basins, in P. A. Allen Devonian events and correlations: Geological Society and P. Homewood, eds., Foreland basins: Special Publica- (London) Special Publication 278, p. 39–81. tions of the International Society of Sedimentologists, v. 8, Ver Straeten, C. A., and C. E. Brett, 1995, Lower and Middle p. 369–392. Devonian foreland basin fill in the Catskill front: Strati- Tankard, A. J., 1986b, Depositional response to foreland de- graphic synthesis, sequence stratigraphy, and the Aca- formationintheCarboniferousofeasternKentucky: dian Orogeny: New York State Geological Association, AAPG Bulletin, v. 70, p. 853–868. 67th Annual Meeting Guidebook, p. 313–356. Thomas, W. A., 1977, Evolution of the Appalachian-Oauchita Ver Straeten, C. A., and C. E. Brett, 2000, Bulge migration salients and recesses from reentrants and promontories and pinnacle reef development, Devonian Appalachian in the continental margin: American Journal of Science, foreland basin: Journal of Geology, v. 108, p. 339–352, v. 277, p. 1233–1278. doi:10.1086/314402. Turcotte, D. L., and G. Schubert, 1982, Geodynamics: Ap- Ver Straeten, C. A., and C. E. Brett, 2006, Pragian to Eifelian plications of continuum physics to geological problems: strata (middle Lower to lower Middle Devonian), north- New York, John Wiley, 450 p. ern Appalachian Basin: Stratigraphic nomenclatural USGS (U.S. Geological Survey) National Geologic Map changes: Northeastern Geology and Environmental Database Geologic Names Lexicon, 2008, http:ngmdb Sciences, v. 28, p. 80–95. .usgs.gov/Geolex/geolex_home.html (accessed August Ver Straeten, C. A., D. H. Griffing, and C. E. Brett, 1994, 2010). The lower part of the Middle Devonian Marcellus Vail, P. R., 1987, Seismic stratigraphy interpretation proce- “Shale,” central to western New York state: Stratigraphy dure, in A.W. Bally, ed., Atlas of seismic stratigraphy, and depositional history: New York State Geological As- vol. 1: AAPG Studies in Geology 27, p. 1–10. sociation, 66th Annual Meeting Guidebook, p. 271–321. Van Tassell, J., 1994, Evidence for orbitally-driven sedimen- Wagner, W. R., 1976, Growth faults in Cambrian and Lower tary cycles in the Devonian Catskill delta complex, in J. Ordovician rocks of western Pennsylvania: AAPG Bul- M. Dennison and F. R. Ettensohn, eds., Tectonic and letin, v. 60, p. 414–427. eustatic controls on sedimentary cycles: Society for Sedi- Werne, J. P., B. B. Sageman, T. W. Lyons, and D. J. Hollander, mentary Geology, Concepts in Sedimentology and Pa- 2002, An integrated assessment of a “type euxinic” de- leontology, v. 4, p. 121–131. posit: Evidence for multiple controls on black shale de- Van Tyne, A., 1983, Natural gas potential of the Devonian position in the middle Devonian Oatka Creek formation: black shales of New York: Northeastern Geology and American Journal of Science, v. 302, p. 110–143, doi:10 Environmental Sciences, v. 5, p. 209–216. .2475/ajs.302.2.110. Van Wagoner, J. C., H. W. Posamentier, R. M. Mitchum, P. R. West, M., 1978, Preliminary stratigraphic cross section show- Vail, J. F. Sarg, T. S. Loutit, and J. Hardenbol, 1988, An ing radioactive zones in the Devonian black shales in the overview of the fundamentals of sequence stratigraphy eastern part of the Appalachian Basin: U.S. Geological and key definitions, in C.K.Wilgus,B.S.Hastings, Survey, Oil and Gas Investigations Chart OC-86. C. G. St. C. Kendall, H. Posamentier, C. A. Ross, and Wheeler, R. L., 1980, Cross-strike structural discontinuities: J. Van Wagoner, eds., Sea-level changes: An integrated Possible exploration tool for natural gas in Appalachian approach: SEPM Special Publication 42, p. 39–45. overthrusting belt: AAPG Bulletin, v. 64, p. 2166–2178. Van Wagoner, J. C., R. M. Mitchum, K. M. Campion, and Williams, E., and W. Bragonier, 1974, Controls of early Penn- V. D. Rahmanian, 1990, Siliciclastic sequence stratigra- sylvanian sedimentation in western Pennsylvania: Geo- phy in well logs, cores, and outcrops: Concepts for high- logical Society of America Special Paper 148, p. 135–152.

Lash and Engelder 103